Presentation is loading. Please wait.

Presentation is loading. Please wait.

Status of Ecloud Build-Up Simulations for the ILC DR’s

Similar presentations


Presentation on theme: "Status of Ecloud Build-Up Simulations for the ILC DR’s"— Presentation transcript:

1 Status of Ecloud Build-Up Simulations for the ILC DR’s
Miguel A. Furman LBNL ILCDR ecloud mtg. 10 March 2010

2 } Summary Essential simulation input parameters POSINST code features
Results: DC04 and DSB3 peak SEY: dmax=0, 0.9, 1.0, ..., 1.4 field-free region and dipole bend with and without antechamber Conclusions Results seem consistent with Theo Demma’s (15dec09), although an apples-to-apples comparison remains to be carried out THE FINE PRINT: this is work in progress. The results presented here are based on one set of input parameters, albeit believed to be realistic. Computational parameters have been only partially exercised to establish numerical stability. } in all combinations

3 Simulation input parameters for DC04 & DSB3 (mostly from M
Simulation input parameters for DC04 & DSB3 (mostly from M. Pivi, 17 Nov et. seq.) Beam energy Eb=5 GeV Bunch population Nb=2x1010 RMS bunch length sz=5 mm Bunch train 45 bunches (spacing tb = ns = 4 buckets) Gap length between trains 15x4=60 buckets Fill pattern simulated 5 x (train+gap) Chamber radius a=2.5 cm Antechamber full height (if present) h=1 cm Antechamber clearing efficiency h=98% Quantum efficiency of chamber surface QE=0.1 Radiation vertical spot size at wall sy=1 mm Photon reflectivity R=0.9 (*) Peak SEY values explored dmax=0, 0.9, 1.0, 1.1, 1.2, 1.3,1.4 Electron energy at dmax Emax=296 eV SEY at E=0 d(0)=0.31xdmax (*) This implies that, if there is no antechamber, a fraction 1–R=0.1 of the photoelectrons are generated localized at the right “edge” of the chamber. If there is an antechamber, a fraction 1–R=5.5x10–8 of the photoelectrons are generated localized at the right “edge” of the chamber (just above and below the antechamber opening).

4 Input parameters that vary from DC04 to DSB3
Circumference [m] 6476.4 3238.2 Harmonic number 14042 7021 ng [photons/e+/m] (radiated g’s) 0.33 0.47 ne [photo.-el./e+/m] (without antech.) 0.033 0.047 (with antech.) 0.66x10–3 0.94x10–3 field-free bend (sx,sy) [mm] (360,6) (260,6) (270,6) (110,5) B [T] 0.27 0.36 NB: ne = ng x (QE) x (1–h), where h=0.98 is the antechamber clearing efficiency

5 Computational parameters for all cases
Bunch profile 3D gaussian Full bunch length 5sz Integration time step during bunch Dt = 1.25x10–11 s (= 9 kicks/bunch) Integration time step if no bunch present Dt = (2.4–2.5)x10–11 s Space-charge grid 64x64 Grid cell size (50 mm)/64=781 mm Macro-photoelectrons per bunch passage 1,000 Max. number of macroparticles allowed 20,000

6 SEY components Based on TiN fits (M. Pivi) Explored dmax=0,0.9–1.4
keeping Emax=296 eV fixed while scaling d(0)0.31 x dmax NB: when changing dmax away from dmax=1, scale all 3 components (TS, R, E) by the same factor realism of this scaling is subject to debate

7 Chamber cross-section without/with antechamber opening

8 “POSINST” code build-up simulations
Simulate individual sections of the ring, one at a time Field-free or dipole bend Round pipe, a=2.5 cm, with/without antechamber of FH=1 cm Compute instantaneous and average ecloud density and many other quantities over 5 trains of 45 bunches each this is long enough for sensible time averages Use actual values for Nb, sx, sy, sz Use actual chamber geometry

9 Results Build up Time-averages vs dmax
Aver. density (time and space) Density in front of bunch within the 10-s beam ellipse NB: “front of bunch” is defined to be Dz=2.5sz from center Aver. beam neutralization Everything else that POSINST computes (not shown here) is available by request

10 Field-free region build-up space-averaged ecloud density
DC04, w/o antch. DC04, w antch. DSB3, w antch. DSB3, w/o antch.

11 Bending magnet build-up space-averaged ecloud density
DC04, w/o antch. DC04, w antch. DSB3, w antch. DSB3, w/o antch.

12 Time averages over 5 trains: overall density and 10-s front-bunch density NB1: aver. ne  0.7 x (peak ne). NB2: 10-s ne is very noisy w.r.t. aver. aver. ne aver. ne 10-s ne 10-s ne

13 Aver. beam neutralization
DSB3 DC04 Neutralization is significant if there is no antechamber

14 Overall ne at saturation(*) units: 1012 m–3
DC04 DSB3 field-free bend dmax antch. no antch 0.031 1.5 0.032 1.4 0.044 2.2 0.045 1.8 0.9 0.056 3.0 0.054 0.081 4.3 0.090 3.3 1.0 0.064 3.4 0.058 2.4 0.092 4.6 0.10 3.7 1.1 0.073 3.9 0.065 2.8 5.3 0.12 1.2 0.087 4.7 0.079 3.2 6.0 0.16 5.1 1.3 5.4 0.11 4.1 0.15 6.6 >0.2 6.1 0.14 6.3 >0.8 5.0 0.20 7.3 >1 7.0 (*) “Saturation” means here: “at the end of the last (5th) train of bunches”

15 ne within 10 beam s’s at saturation(*) units: 1012 m–3
DC04 DSB3 field-free bend dmax antch. no antch 0.08 5.0 0.01 0.6 0.12 9 0.015 0.7 0.9 0.18 10 0.035 1.6 0.22 14 0.03 1.5 1.0 0.20 11 0.046 0.26 0.04 2.0 1.1 0.065 3.1 0.31 19 0.09 2.3 1.2 0.25 15 0.11 4.5 0.41 20 0.05 3.0 1.3 0.35 16 6.0 0.48 23 0.2 3.5 1.4 0.44 >4 8.0 0.62 24 >0.6 (*) “Saturation” means here: “at the end of the last (5th) train of bunches.” NB.: these data typically have large statistical errors, ~50%.

16 ne at bunch front within 10 beam s’s (*) units: 1012 m–3
DC04 DSB3 field-free bend dmax antch. no antch 0.024 1.2 0.023 1.0 0.034 1.7 0.031 1.3 0.9 0.044 2.3 0.038 1.6 0.063 3.2 2.4 0.050 2.6 0.042 1.8 0.070 3.6 0.073 1.1 0.057 3.0 0.048 1.9 0.081 4.0 0.086 2.9 0.066 3.4 0.056 2.2 0.94 4.5 0.10 0.080 3.9 0.079 0.11 5.0 >0.2 1.4 >0.3 3.1 0.14 5.6 4.6 (*) Note: these simulated data have large errors (~30-40%) due to statistical noise. Within these errors, there is no difference between the time-averaged density and the instantaneous density at the last bunch in the train

17 Aver. e–-wall impact energy <E0>
DC04 DSB3 Significantly below Emax=296 eV Tentative predictions: If all else is fixed, ecloud density will be higher if Emax is lower than 296 eV, and viceversa Ditto if Nb is larger than 2x1010 Why does <E0> depend strongly on dmax in some cases?

18 Conclusions Significant no. of photoelectrons in all cases
~20% (or more) of the ecloud density, depending on dmax Amplification effect from secondary emission is a factor ~ a few This is very small, in my experience (?) ne: generally smooth, monotonic dependence on dmax in the range examined Exception: ne has a 1st-order phase transition in bends with antech. at dmax≈ ecloud density in DSB3 is larger than in DC04 by 10–20% Antechamber: in field-free regions: ecloud density lower by factor ~30-40 relative to no antechamber for all cases explored in bending magnets: ecloud density lower by factor ~30-40 relative to no antechamber unless dmax exceeds ~1.3 (DC04) or ~1.2 (DSB3), in which case these results are inconclusive but it looks looks to me like antechamber won’t provide much protection, if any, in these cases Aver. density: with ant.: ne~1x1011 m–3; without ant.: ne~(2-4)x1012 m–3 Aver. beam neutralization: with ant.: ~1-2%; without ant.: % 10-s front bunch density comparable to aver. density within a factor of less than 2 For DC04 dipole with antechamber and dmax=1.2, ne≈1x1011 m–3, consistent with T. Demma’s result (15dec09); but he used R=0.5 and assumed h=97%

19 Caveats Numerical convergence partly checked
If Dt3Dt, results do not change much, except for bends with antechamber and large dmax Dependence on space-charge grid not checked But 64x64 has given quite stable results in other cases Ditto for no. of macroparticles Reflectivity parameter R not exercised But high values (like R=0.9, used in all cases here) tends to yield higher ne than low R, especially for bends Sensitivity to details of SEY not explored, except for dmax It seems desirable to at least vary Emax by ±20% and see what happens Ditto for the SEY relative composition TS/R/E Not done: quads, wigglers, and other regions of the machine


Download ppt "Status of Ecloud Build-Up Simulations for the ILC DR’s"

Similar presentations


Ads by Google